|Publication number||US6459910 B1|
|Application number||US 08/481,972|
|Publication date||Oct 1, 2002|
|Filing date||Jun 7, 1995|
|Priority date||Jun 7, 1995|
|Publication number||08481972, 481972, US 6459910 B1, US 6459910B1, US-B1-6459910, US6459910 B1, US6459910B1|
|Inventors||Theodore W. Houston|
|Original Assignee||Texas Instruments Incorporated|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (13), Referenced by (32), Classifications (15), Legal Events (5)|
|External Links: USPTO, USPTO Assignment, Espacenet|
1. Field of the Invention
This invention relates to the use of a speech compression technique to reduce the amount of information required to enable reasonably accurate reproduction of the original speech such that it can be transmitted at low bandwidth for pager or low power mobile telephone systems.
2. Brief Description of the Prior Art
Pager systems use narrow bandwidth to transmit a phone number to a portable unit. When the portable unit receives the number, the user must go to a telephone to either receive a recorded message or to call the initiating number. Alternatively, the initiating caller can send a keyed in message. It would be desirable for the caller to provide a spoken message that would be transmitted to the pager, but that would require a wider bandwidth than is generally used by pager systems. It would further be desirable to be able to use a pager system to send a spoken reply message, but that again would require more bandwidth and would also require considerably more power in the portable unit.
There is a desire to have mobile phones operate for longer time periods on smaller batteries. One solution is to have a high density of relay stations so that low power transmission can be used. However, this will not always be possible and a high power transmission may sometimes be required to reach the nearest relay station. Thus, it would be desirable to have a system that reduces required transmission power by reducing the amount of data that is needed for conveying a message.
Speech recognition systems have now reached the degree of sophistication whereby they are capable of recognizing and identifying many spoken words and phrases or phonemes and providing digital codes for representation, storage, transmission and reproduction of such words or phrases and/or phonemes. Such digital code greatly compresses the original audio information, but retains sufficient information to allow recreation of the original speech in an understandable form. Word or phrase recognition provides the greatest compression while phoneme recognition is more flexible.
Transmission of analog information, particularly sound, generally requires a relatively large amount of power and relatively large bandwidth. It is readily apparent that, with the bandwidth restrictions being placed upon transmission channels, the amount of information required to be transmitted within the allowed limited channel bandwidth in order to provide the same or similar sound recognition information at the receiving end will require additional transmission time. While digital techniques require less bandwidth than do analog techniques, the amount of digital information required for accurate duplication of the analog signal is considerably more than required for the identification of words or phonemes. For example, a few bits can represent a word or phoneme, while several bits may be used to represent a fraction of a second of analog information, where the number of bits and the duration of time represented depends upon the accuracy of reproduction to be achieved. Since the power utilized in transmission of information or data is far greater than the power generally required to manipulate the information or data to be transmitted, it is apparent that any procedures that will minimize transmission power requirements will be extremely useful, especially when battery operated devices with limited power availability are involved. Bandwidth restrictions are also important. This minimization of transmission power requirements becomes extremely important in devices such as pagers and mobile telephones which are battery operated and have limited available power. It is apparent that the amount of information that can be received and/or transmitted by the pager or mobile telephone will have a direct relation to the amount of available power and bandwidth and this amount of received information can be increased if it can be transmitted more efficiently and the ratio of message information transmitted to power consumption and bandwidth required can be increased.
While the discussion herein is primarily directed to spoken input, it can also be applied to typed input. An abbreviated code can be assigned to words and phrases and this abbreviated code can be substituted for the input string of characters for storage and transmission.
In accordance with the present invention, information, primarily but not limited to speech, is compressed to reduce the amount of transmitted information required to enable intelligible reproduction of the original information at a remote location.
In the preferred embodiment, the information is compressed at a pager base station prior to transmission to the remote pager. The remote pager stores the transmitted compressed information and recreates the information from the stored code on command.
Briefly, the above compression is accomplished by recognizing certain information, such as phonemes and/or words or phrases and/or any other types of information capable of being identified by a relatively small amount of digital data. The recognizable information capable of being identified with a small amount of data is coded at the transmitting station by comparison with a data base and providing the digital coding for transmission. In addition, the type of voice is recognized, such as male, female, child, foreign accent or a particular individual and coding is sent indicating such voice type along with the other recognizable information, such as volume or pitch. The amount of information detected and transmitted relative to voice type will determine the accuracy with which the voice is reproduced at the remote station. However, though an increased amount of information will provide more accurate voice reproduction, it will also require that an increased amount of information be transmitted, thereby utilizing more data transmission time. Also, additional processing of the information will be required both for compression and recreation. It follows that a tradeoff must be made. However, by transmitting voice recognition information as well as information relative to phonemes, words, phrases and the like, a more accurate reproduction of the original voice can be provided with a greatly reduced amount of transmitted information. For example, if the voice characteristics of certain individuals are also coded in the data base at each terminal, a single code relative to the individual talking can be provided, thereby providing an accurate voice reproduction at the receiving end while only one voice identifier and phoneme data are being transmitted.
Optionally, users of the system can have voice characteristics stored in the system data base and input an assigned code (for example, saying their name) to select their voice characteristics for the speech recreation.
The coding is transmitted and the transmitted code is identified with a corresponding code at the receiving end (e.g., at the central station and the remote pager). When the particular information capable of being compressed in the manner described herein is recognized at the transmission end by comparison in a data base and the type of voice or individual speaking is also recognized by comparison in a data base, the particular binary codes in the data base for that information are transmitted to the remote unit (e.g., the pager or mobile telephone).
The coded information to describe the type of information sent is preferably binary, thereby requiring very limited bandwidth for transmission. Also, the binary number representative of the information capable of being compressed can be transmitted to the remote location using much less power since the amount of data to be transmitted is substantially reduced. This coded information received at the remote location, such as a pager or mobile telephone, is used to reconstruct the information capable of being compressed as well as the voice type from the data base therein which has the same coded information as in the data base at the transmitting end. As stated above, in the case of a spoken message, the transmitted data can provide additional information such as, for example, whether the voice was male, female, a youth, accented, a particular individual, etc. in order to reproduce the original sound more closely. The remote station can also have some or all of the operational capabilities of the transmitting station, thereby permitting at least some type of communication in both directions. For mobile telephone applications, the preferred form is to have the mobile unit receive information in the standard format and to transmit information in the encoded format to save transmission power. The base station would optionally recreate audio information (either analog or digital) from the encoded information for further transmission in the telephone network. Optionally, the mobile unit could switch from standard transmission to encoded transmission, depending upon battery status and distance from the receiving station.
The representation of the phoneme or other information can be altered to the point that it is still recognizable, be close to the sound of the originally spoken information and yet can be transmitted using much less information than that required to transmit the phoneme as originally expressed.
The FIGURE is a schematic diagram of a paging system in accordance with the present invention.
Referring now to the FIGURE, there is shown a pager system having a base transmitting section 1 and a receiving section 3 in the form of a pager. It should be understood that the transmitting section 1 can also be a receiving section and the receiving section 3 can also be a transmitting section to provide a two-way pager if sufficient power capability is available at the pager section 3.
Speech which requires a relatively large bandwidth for accurate reproduction is generated by a voice or the like at the transmitting section 1 and is sensed by an encoder 5 which includes a data base represented by way of example in the FIGURE by a phoneme recognition unit 7 of standard type having its own data base and a voice type recognition unit 9 having its own data base. The phonemes detected by phoneme recognition unit 7 require a relatively large bandwidth for accurate transmission are compared with the phoneme data base which matches the recognized phoneme and provides a binary number corresponding to the encoder in standard manner. The voice type unit 9 determines the type of voice or a specific person's voice by recognition of pitch and harmonic content, or by input code, and by comparison with the data base therein and provides a binary number corresponding to the voice type information received. The binary numbers corresponding to the detected phoneme requires substantially less data than the original analog signal or digital representation of the analog signal for accurate transmission of the phonemes themselves and therefore less transmission power and less bandwidth are required than does the phoneme itself as originally applied to the data base 9. The amount of binary data required to provide voice recognition is also small with the amount of data required to transmit both phoneme and voice recognition data being far less than that required to provide transmission of the original analog equivalent.
The binary information indicative of voice type and phonemes detected at encoder 5 is then transmitted by a transmitter 13, at low bandwidth and relatively low power requirement compared with that required if the phoneme in analog form or digital representation of the analog form were transmitted, to the receiver 13 of the receiving station 3. The receiving station 3 is a pager or mobile telephone unit in the preferred embodiment. The receiver 13 receives the binary signals and stores these signals for later use in storage 15 or operates upon these signals concurrently with reception. Upon conditioning by the user, the pager then recreates the analog or digital representation of the sound in conjunction with a data base. This is accomplishes by selecting a message from the stored code 15 with a code selector 17 where the storage option is provided and utilized and passing this code via the recreation circuit 19 to the code data base 21. The code data base 21 recreates the phonemes and voice type by comparison of the code with code in the data base and returning the recreated information to the recreation circuit. The recreation circuit 19 determines whether the message is aural or visual. If the message is aural, then the phonemes with voice type are provided in aural form by the broadcast and display device 23 with original voice type reproduced to the extent coded. If the message is visual, then the message is displayed on a display in device 23 in standard manner.
Upon receipt of information at the receiver 13, a signal can be provided by the pager in standard manner to indicate such reception.
The receiving station 3 requires relatively low power operation since it receives a binary representation addressed to more complex information stored in its code data base 21. Sections 19, 21 and 23 can be powered only as required when the user elects to hear a message.
In the above described embodiment, the amount of information contained in the binary number transmitted can be varied, as desired, this amount generally being a tradeoff among reproduction quality, transmission time and circuit complexity.
In addition, it is apparent that the circuitry shown can be used to transmit from the receiving unit 3 back to the receiving unit 1 in the same manner as demonstrated above from the transmitting unit 1 to the receiving unit 3. Alternatively, encoding and decoding can both be accomplished at mobile units, optionally with a base station and normal telephone networks for relaying the encoded information. In another alternative, the base station broadcasts selected encoded messages.
Though the invention has been described with reference to specific preferred embodiments thereof, many variations and modification will immediately become apparent to those skilled in the art. It is therefore the intention that the appended claims be interpreted as broadly as possible in view of the prior art to include all such variations and modifications.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4408099 *||Jun 5, 1981||Oct 4, 1983||Nippon Electric Co., Ltd.||Radio paging system capable of accepting message with access of paging receiver|
|US5029200 *||May 2, 1989||Jul 2, 1991||At&T Bell Laboratories||Voice message system using synthetic speech|
|US5086394 *||Mar 29, 1990||Feb 4, 1992||Shmuel Shapira||Introduction system for locating compatible persons|
|US5122795 *||Mar 1, 1989||Jun 16, 1992||Metrocast||Scanning receiver for nationwide radio paging system|
|US5274495 *||Sep 17, 1991||Dec 28, 1993||Fujitsu Limited||Optical amplifier|
|US5297183 *||Apr 13, 1992||Mar 22, 1994||Vcs Industries, Inc.||Speech recognition system for electronic switches in a cellular telephone or personal communication network|
|US5351276 *||Feb 10, 1992||Sep 27, 1994||Simpact Associates, Inc.||Digital/audio interactive communication network|
|US5444673 *||Jul 12, 1994||Aug 22, 1995||Mathurin; Trevor S.||Audio controlled and activated wristwatch memory aid device|
|US5481255 *||Feb 10, 1993||Jan 2, 1996||Data Critical Corp.||Paging transmission system|
|US5511111 *||Nov 1, 1993||Apr 23, 1996||Engineering And Business Systems, Inc.||Caller name and identification communication system with caller screening option|
|US5528728 *||Jul 12, 1993||Jun 18, 1996||Kabushiki Kaisha Meidensha||Speaker independent speech recognition system and method using neural network and DTW matching technique|
|US5592531 *||Apr 3, 1995||Jan 7, 1997||Motorola, Inc.||Selective call communication system having adaptive voice data entry|
|US5652789 *||Sep 30, 1994||Jul 29, 1997||Wildfire Communications, Inc.||Network based knowledgeable assistant|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US6771982 *||Oct 12, 2000||Aug 3, 2004||Curo Interactive Incorporated||Single action audio prompt interface utlizing binary state time domain multiple selection protocol|
|US6804539 *||Sep 19, 2002||Oct 12, 2004||Curo Interactive Incorporated||Single action audio prompt interface utilizing binary state time domain multiple selection protocol|
|US7010314 *||May 17, 1999||Mar 7, 2006||Jean-Francois Taillens||Method for remote connection of machines and corresponding electronic analog apparatus|
|US7437287 *||Nov 1, 2004||Oct 14, 2008||Samsung Electronics Co., Ltd||Apparatus and method for voice recognition and displaying of characters in mobile telecommunication system|
|US7475343 *||May 11, 1999||Jan 6, 2009||Mielenhausen Thomas C||Data processing apparatus and method for converting words to abbreviations, converting abbreviations to words, and selecting abbreviations for insertion into text|
|US7668567 *||Jan 21, 2005||Feb 23, 2010||Toupin Paul M||Single action audio prompt interface utilising binary state time domain multiple selection protocol|
|US7801508 *||Nov 12, 2004||Sep 21, 2010||Voicecash Ip Gmbh||Method for authentication of a user on the basis of his/her voice profile|
|US8090410||Jul 23, 2010||Jan 3, 2012||Voicecash Ip Gmbh||Method for authentication of a user on the basis of his/her voice profile|
|US8098600||Feb 1, 2010||Jan 17, 2012||Parus Holdings, Inc.||Computer, internet and telecommunications based network|
|US8155708||Feb 22, 2010||Apr 10, 2012||Curo Interactive Incorporated||Single action audio prompt interface utilising binary state time domain multiple selection protocol|
|US8611955 *||Apr 9, 2012||Dec 17, 2013||Curo Interactive Incorporated||Single action audio interface utilising binary state time domain multiple selection protocol|
|US8838074||Mar 4, 2013||Sep 16, 2014||Parus Holdings, Inc.||Computer, internet and telecommunications based network|
|US8843120||Jan 13, 2012||Sep 23, 2014||Parus Holdings, Inc.||Computer, internet and telecommunications based network|
|US8843141||Jul 17, 2013||Sep 23, 2014||Parus Holdings, Inc.||Computer, internet and telecommunications based network|
|US9232037||Dec 17, 2013||Jan 5, 2016||Curo Interactive Incorporated||Single action sensory prompt interface utilising binary state time domain selection protocol|
|US9377992||May 26, 2010||Jun 28, 2016||Parus Holdings, Inc.||Personal voice-based information retrieval system|
|US9451084||May 3, 2012||Sep 20, 2016||Parus Holdings, Inc.||Robust voice browser system and voice activated device controller|
|US9571445 *||Jun 29, 2007||Feb 14, 2017||Parus Holdings, Inc.||Unified messaging system and method with integrated communication applications and interactive voice recognition|
|US9734819 *||Feb 21, 2013||Aug 15, 2017||Google Technology Holdings LLC||Recognizing accented speech|
|US9769314||Jun 27, 2016||Sep 19, 2017||Parus Holdings, Inc.||Personal voice-based information retrieval system|
|US20020046030 *||May 16, 2001||Apr 18, 2002||Haritsa Jayant Ramaswamy||Method and apparatus for improved call handling and service based on caller's demographic information|
|US20020077811 *||Dec 14, 2001||Jun 20, 2002||Jens Koenig||Locally distributed speech recognition system and method of its opration|
|US20030017847 *||Sep 19, 2002||Jan 23, 2003||Toupin Paul M.||Single action audio prompt interface utilising binary state time domain multiple selection protocol|
|US20030023431 *||Jul 26, 2001||Jan 30, 2003||Marc Neuberger||Method and system for augmenting grammars in distributed voice browsing|
|US20050065791 *||Nov 1, 2004||Mar 24, 2005||Samsung Electronics Co., Ltd.||Apparatus and method for voice recognition and displaying of characters in mobile telecommunication system|
|US20050107070 *||Nov 12, 2004||May 19, 2005||Hermann Geupel||Method for authentication of a user on the basis of his/her voice profile|
|US20050124385 *||Jan 21, 2005||Jun 9, 2005||Toupin Paul M.||Single action audio prompt interface utilising binary state time domain multiple selection protocol|
|US20070263601 *||Jun 29, 2007||Nov 15, 2007||Parus Interactive Holdings||Computer, internet and telecommunications based network|
|US20100153117 *||Feb 22, 2010||Jun 17, 2010||Curo Interactive Incorporated||Single action audio prompt interface utilising binary state time domain multiple selection protocol|
|US20100291901 *||Jul 23, 2010||Nov 18, 2010||Voicecash Ip Gmbh||Method for authentication of a user on the basis of his/her voice profile|
|US20120198340 *||Apr 9, 2012||Aug 2, 2012||Curo Interactive Incorporated||Single action audio interface utilising binary state time domain multiple selection protocol|
|US20140236595 *||Feb 21, 2013||Aug 21, 2014||Motorola Mobility Llc||Recognizing accented speech|
|U.S. Classification||455/563, 379/88.03, 379/88.13, 455/422.1, 704/E19.007, 455/550.1, 455/458, 379/88.01, 379/88.04, 379/88.02|
|International Classification||G10L19/00, H04W88/02|
|Cooperative Classification||H04W88/023, G10L19/0018|
|Jun 7, 1995||AS||Assignment|
Owner name: TEXAS INSTRUMENTS INCORPORATED, TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:HOUSTON, THEODORE W.;REEL/FRAME:007552/0038
Effective date: 19950606
|Mar 28, 2006||FPAY||Fee payment|
Year of fee payment: 4
|Mar 23, 2010||FPAY||Fee payment|
Year of fee payment: 8
|Mar 26, 2014||FPAY||Fee payment|
Year of fee payment: 12
|Jan 19, 2017||AS||Assignment|
Owner name: INTEL CORPORATION, CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:TEXAS INSTRUMENTS INCORPORATED;REEL/FRAME:041383/0040
Effective date: 20161223